jaredsohn

In today's AI times, I find it a little amusing to think about emojis as an automation of the craft of making ascii art. Is a little different since people don't get paid for that, but there was a creative component to it.

I completely agree that unverified claims create a heavy burden for maintainers. My only point was about the language used: 'disparaging' to me implies a bad-faith attack or a dismissive attitude, whereas this was just an honest technical mix-up that the poster immediately corrected.

I think part of the confusion with that word comes from things like corporate non-disparagement clauses. In those contracts, lawyers write the terms so broad that "disparagement" means saying anything negative, regardless of malice or intent.

FWIW, I don't think the GP post is disparaging (at least as I read it right now.)

I think it is fair to list limitations from using a library that provides an abstraction; it can suggest why a tool isn't right for a person's use cases.

But it also sounds like this API handles those pretty well.

I learned about Tile via ads in its early days and then eventually bought it.

Maybe one of those trillionaires could help for a bit before leaving to make his own AI model, too.

I think a reason for this is suppose the next year you run into some difficulties so it requires 14.2M. Now you have to fight to request an extra 0.2M added to your budget that you wouldn't have to worry about if you had 15M.

'Everyone over 30 knows this' is a prior assumption (it is not necessarily correct; and nothing is said about shame).

The comic strip is saying if above is true, then people still have to learn at some point so on average it would be around 10k people per day.

I think the math is this:

For people born in a given year: 4000000/365/30 = 365 people per day

but you have 30 sets of those people (those born this year, those born last year, those born two years ago, etc.) So 365 * 30 = 10950. 10k is easier to say for viral purposes.

Reminds me of ZeptoLab (company that made Cut The Rope)'s Pudding Monsters but it adds a lot more variety. (And it combines them together within one square instead of multiple squares and they don't fall off the edge of the screen.)

Models like Haiku are great if you just want to apply JSON structure to some free text.

For SMS, you can block text messages that match keywords via apps like VeroSMS on the iPhone. There may be similar apps on Android.

I personally don't think it is possible to get yourself off all of the lists so I do that and mark as spam in gmail.

Strangely I think it was TechCrunch. But I was reading pg's essays years earlier (maybe through Joel On Software) and already had known about yc.

Think it was related to the initial request for startups post: https://techcrunch.com/2009/08/16/y-combinator-starts-seedin...

This posting made it 15 :)

Might be date-related; all are from past 3 days

The site has been updated now to use the current domain.

Weird thing is that domain is even available to buy.

If you do have AI write for you, I think it is useful to indicate what is going on by having a rule set up for it. That can at least prepare the reader for reading AI text.

I personally find it okay / convenient to have AI respond to PR review comments with respect to something being addressed or why it was not. That text is often pretty mechanical.

day or 2 max

I've frequently seen tasks that it thinks will take weeks being done in under an hour. And it will often recommend doing X instead of Y because X requires so much extra work. Basically I just remind it that it is an LLM.

If it worries something is error prone, I ask it to write tools to verify it.

The line numbers were also used for GOTO and GOSUB (although later on languages supported strings instead of numbers).

There are tools that use LLMs to do this.

I've done this manually by building a big feature branch and asking an LLM to extract out functionality for a portion of it.

For the former, it would seem to split based on frontend/backend, etc. rather than what semantically makes the most sense and for the latter it would include changes I don't want and forget some I do want. But I haven't tried this a lot.

The GP said more with LLMs than people - not no interactions at all with people and not preferring machines to people. I don't think it is that hard to spend more time talking with LLMs than people if you work in tech and I don't think that takes away from one's life meaningfulness.

One thing that is useful to remember is that if you ask AI for help on using some app, it will likely refer to the mobile UI instead of the web UI. I find it annoying that sometimes there are features that are only available in the mobile UI.

Interestingly, they show up in the company list. When you click the link it returns 404.

https://www.ycombinator.com/companies/?query=delve

that conversation is lost forever.

You should be able to find it in ~/.claude

You can also ask Claude to search your history to answer questions about it.

Your original post (and even after this comment I think) was vague in that AI can be used in a lot of different ways in 'production' - to generate code, to manage deployment / scripts, or as part of a feature that uses inference.

For example, if you're writing code with AI, you can still review it just like you would if a colleague wrote it. You can write tests (or have the AI do so) to prevent some hallucinations, too.

This question is too vague to answer.

This xkcd is again relevant: https://xkcd.com/2501/

I find it great for nutrition and cooking. I used AI to build a food tracker that works at the micronutrient level (LLMs built the code but are mostly not used in the app itself) but many of my prompts are helping me to learn nutrition in general.

I also use it to help me learn to cook more/better so I ask it questions about the how / why and have made a database of recipes that I tweak based on interests / abilities.

A neat thing is you can always ask AI - what can I do to eat or cook better? Then you do it and keep repeating.

Treat context like git shas. Yes, there is a specific order within a 'branch' but you should be able to do the equivalent of cherry-picking and rebasing it

I connected a few times today to the IPv4 one. Have had no problems myself.

To watch Star Wars in ASCII.

telnet towel.blinkenlights.nl https://www.youtube.com/watch?v=Mhcf6tc2jeQ

(Remember hearing about this a long time ago (from some searching I think it was in 1999 via Slashdot) and verified some instance of it still exists/works.)

To focus on Excel / Google Sheets itself, AI makes it a lot more usable in that you can describe what you want and it can generate formulas for you. I have a running spreadsheet that I had to manually manipulate because I didn't know how to automate some parts and a few minutes with an LLM fixed it for me.